AI Product Manager | Engineer by Training | Applied AI & Data Science for Subscriptions & Growth

Hi 👋
I’m a product leader with an engineering background, focused on applying
AI and Data Science as product capabilities to solve subscription, growth, and lifecycle problems in consumer products.

I don’t build models for the sake of accuracy.
I build decision-support systems that help product teams answer questions like:

• Why are users not converting?
• Which behaviors actually signal intent?
• Where should we intervene — and where should we not?
• What trade-offs are we making when we optimize?

Earlier in my career, I worked as a Technical Evangelist at Microsoft and Nokia and spent several years as an Android developer, building and supporting mobile products across diverse ecosystems.

Today, while I no longer work as a full-time developer, I remain deeply technical. I operate at the intersection of:

• Product strategy
• Data Science & Machine Learning
• Engineering-aware decision making

My focus is on applied AI — turning data, signals, and models into clear, actionable product decisions, not just analytical insights.

An end-to-end case study focused on subscription conversion in a streaming-style product.

This project demonstrates how I:

• Frame ambiguous business problems as decision systems
• Design realistic behavioral signals (not toy features)
• Train interpretable baseline models aligned with product constraints
• Explicitly document assumptions, risks, and trade-offs
• Translate model outputs into product and experimentation decisions

📂 Repository:
🔗 https://github.com/diegogallegof/ai-product-portfolio

👉 Recommended starting point:
01_user_conversion/

This case is intentionally built as a system, progressing from: data generation → modeling → insights → recommendations → service exposure.

This is not a collection of tutorials, toy notebooks, or Kaggle-style projects.

This GitHub is a case-based AI Product portfolio, grounded in real product constraints, such as:

• Limited experimentation bandwidth
• Noisy and imperfect behavioral signals
• Trade-offs between accuracy, interpretability, and usability
• The need to make decisions under uncertainty

Across projects, you’ll find:

• 🧠 Product-first framing before any modeling
• 📊 Lifecycle, funnel, and monetization-oriented signals
• 🤖 Interpretable models used for understanding and decision support
• 🧪 Explicit assumptions, limitations, and risks
• 🧩 Clear separation between insights and recommendations
• 🧱 Artifacts designed to evolve into deployable systems

• AI is a means to better decisions, not an end
• Product context and metrics come before model selection
• Interpretability often matters more than marginal accuracy gains
• Every model embeds assumptions and product risk
• Strong AI products require tight alignment between product, data, and engineering

Current and upcoming case studies focus on:

• User Conversion & Monetization
• Customer & Subscriber Segmentation
• Churn & Retention Modeling
• User Feedback & Sentiment Analysis
• Experimentation & Causal Thinking
• Lightweight ML-powered product services (APIs)

Each case follows a consistent structure:
problem → signals → insights → decisions → system.

• Python (pandas, numpy, scikit-learn)
• Jupyter Notebooks
• Data visualization (matplotlib)
• FastAPI (for model-backed services)
• Git & GitHub
• Markdown documentation

• GitHub: https://github.com/diegogallegof
• LinkedIn: https://www.linkedin.com/in/diegogallegof/

This profile evolves in public as I build, test, and ship AI-powered product decision systems.